It is proposed to carry out an extensive investigation into statistical methodology which have directed relevance to the design and analysis of experimental studies, planned or observational, in environmental health. One class of problems to be investigated concerns the analysis of incomplete response-time (e.g. survival) data obtained from ethiologic research studies to evaluate the effect of various risk factors. In particular it is proposed to develop methodology to estimate and compare disease prevalence data arising from exposed and control groups of laboratory animals in a survival/serial sacrifice experiment. The efficient design of such experiments will also be studied. Another broad class of problems to be considered is that of sequential sampling in medical trials with several competing treatments. The properties of various schemes will be investigated.